I gratefully acknowledge the research assistance of Eduardo Magalhaes. For their comments, I thank Christopher Achen, Giulio Gallarotti, Stephen Krasner, David Lake, Charles Lipson, Timothy McKeown, Helen Milner, Edward Ray, Cheryl Schonhardt, Peverill Squire, two anonymous reviewers, and participants in the University of Chicago's International Political Economy and Security Seminar and UCLA's International Political Economy Seminar.

1. For an excellent critique, see Nelson, Douglas, “Endogenous Tariff Theory: A Critical Survey,” American Journal of Political Science 32 (08 1988), pp. 796–837.CrossRefGoogle Scholar

2. Mayer, Wolfgang, “Endogenous Tariff Formation,” American Economic Review 74 (12 1984), pp. 970–85.Google Scholar

3. Brock, William A. and Magee, Stephen P., “The Economics of Special Interest Politics: The Case of the Tariff,” American Economic Review 68 (05 1979), pp. 246–50.Google Scholar

4. If the interest group is large and its objective is a public good (low rivalry in enjoying the benefits and difficulty in excluding noncontributors), it may find collective lobbying difficult because of the incentive for beneficiaries to free ride.

5. See, for example, Pincus, Jonathan, Pressure Groups and Antebellum Tariffs (New York: Columbia University Press, 1977).Google Scholar

6. See Schattschneider, E. E., Politics, Pressure and the Tariff (New York: Prentice-Hall, 1935)Google Scholar; and Bauer, Raymond A., Pool, Ithiel de Sola, and Dexter, Lewis A.. American Business and Public Policy (Chicago: Aldine, 1963).Google Scholar

7. See Cassing, James, McKeown, Timothy, and Ochs, Jack, “The Political Economy of the Tariff Cycle,” American Political Science Review 80 (09 1986), pp. 843–62CrossRefGoogle Scholar. Any market model must contain both a demand and a supply side. Models that lack explicit mention of either demand or supply are implicitly assuming that the omitted variable is exogenously fixed.

8. Caves, Richard E., “Canada's Tariff Structure,” Canadian Journal of Economics 9 (05 1976), pp. 278–300.CrossRefGoogle Scholar

9. See Messerlin, Patrick A., “The Political Economy of Protectionism: The Bureaucratic Case,” Weltwirtschaftliches Archiv, vol. 117, 1981, pp. 469–96CrossRefGoogle Scholar; and Destler, I. M., Making Foreign Economic Policy (Washington, D.C.: Brookings Institution, 1980).Google Scholar

10. Finger, J. M., Hall, H. Keith, and Nelson, Douglas R., “The Political Economy of Administered Protection,” American Economic Review 72 (06 1982), pp. 452–66.Google Scholar

11. Gourevitch, Peter A., Politics in Hard Times (Ithaca, N.Y.: Cornell University Press, 1986), p. 107.Google Scholar

12. Baldwin, Robert E., The Political Economy of U.S. Import Policy (Cambridge, Mass.: MIT Press, 1985).Google Scholar

13. Coughlin, Cletus C., “Domestic Content Legislation: House Voting and the Economic Theory of Voting,” Economic Enquiry 23 (07 1985), pp. 437–48.CrossRefGoogle Scholar

14. Tosini, Suzanne C. and Tower, Edward, “The Textile Bill of 1985: The Determinants of Congressional Voting Patterns,” Public Choice, vol. 54, 1987, pp. 19–25.CrossRefGoogle Scholar

15. Stanwood, Edward, American Tariff Controversies, vol. 2 (Boston: Houghton Mifflin, 1903), p. 264.Google Scholar

16. Terrill, Tom E., The Tariff, Politics and American Foreign Policy, 1874–1901 (Westport, Conn.: Greenwood Press, 1973), pp. 137–40.Google Scholar

17. Wyman, Roger E., “Wisconsin Ethnic Groups and the Election of 1890,” Wisconsin Magazine of History 51 (Summer 1968), p. 292.Google Scholar

18. Stanwood, , American Tariff Controversies, p. 294.Google Scholar

19. Taussig, Frank W., The Tariff History of the United States, 8th ed. (New York: Putnam, 1931), p. 284.Google Scholar

20. The connection between electoral safety and a legislator's propensity to supply tariffs is discussed in a later section of this article.

21. See Wilson, Rick K., “An Empirical Test of Preferences for Pork Barrel: District Level Appropriations for River and Harbor Legislation, 1889–1913,” American Journal of Political Science 30 (11 1986), p. 735CrossRefGoogle Scholar. For a discussion of reasons for longer tenure, see Kernell, Samuel, “Toward Understanding 19th Century Congressional Careers: Ambition, Competition, and Resolution,” American Journal of Political Science 21 (11 1977), pp. 669–93CrossRefGoogle Scholar; and Polsby, Nelson W., “The Institutionalization of the U.S. House of Representatives,” American Political Science Review 62 (03 1968), pp. 144–68.CrossRefGoogle Scholar

22. The period under study has one characteristic of representation that is less useful to me: a high degree of party cohesion, which makes it difficult to distinguish party and constituency factors in the roll-call vote. However, insofar as the partisanship of the period is a result of a high variance in the industrial composition of districts, it makes my technique of analysis easier by allowing geographically flexible tariffs (see footnote 60, below). For a discussion of the period, see Brady, David and Althoff, Phillip, “Party Voting in the U.S. House of Representatives, 1890–1910: Elements of a Responsible Party System,” Journal of Politics 36 (08 1974), p. 760.CrossRefGoogle Scholar

23. See Rogowski, Ronald, “Trade and the Variety of Democratic Institutions,” International Organization 41 (Spring 1987), p. 204.CrossRefGoogle Scholar

24. Hansen, John Mark, “Taxation and the Political Economy of Tariffs,” paper presented at the annual meeting of the American Political Science Association,Washington, D.C., 1988.Google Scholar

25. Terrill, , The Tariff, Politics and American Foreign Policy, p. 9.Google Scholar

26. Davis, Lance E. and North, Douglass C., Institutional Change and American Economic Growth (New York: Cambridge University Press, 1971), p. 181.CrossRefGoogle Scholar

27. See Lebergott, Stanley, The Americans: An Economic Record (New York: Norton, 1984), pp. 150–52Google Scholar. The effective rate of protection is the nominal tariff on a commodity, weighted by the penalty incurred by producers of that commodity if they import inputs for which a tariff must be paid. Higher tariffs on inputs will lower the producer's effective rate of protection.

28. Hawke, G. R., “The United States Tariff and Industrial Protection in the Late Nineteenth Century,” Economic History Review 28 (02 1975), pp. 84–99.CrossRefGoogle Scholar

29. Taussig indicates that the increases in the tariffs on raw wool were “too small to have any serious effect.” The raw sugar tariff was repealed but replaced with a bounty; this may have been of some import in the roll-call vote (see discussion of equation 9). According to Stanwood, the benefits of tariff increases on barley, eggs, meat, potatoes, and butter (benefits that were only large in the case of barley) were confined to the Canadian border area. See Taussig, , The Tariff History of the United States, pp. 257–59Google Scholar; and Stanwood, , American Tariff Controversies, p. 266.Google Scholar

30. The Sherman Silver Purchase Act of 1890 was an attempt to buy the support of farmers and Western mining interest groups by obliging the treasury to purchase silver and mint coins and thereby inflate the money supply. The devaluation of farm debt through inflation would have taken some time to occur and thus provided little short-term benefit to the Republicans. The immediate effect was financial panic and heavy buying of gold, which probably hurt the Republicans.

31. See Baack, Bennett D. and Ray, Edward J., “Special Interests and the Adoption of the Income Tax in the United States,” Journal of Economic History 45 (09 1985), pp. 607–25.CrossRefGoogle Scholar

32. Ashley, Percy, Modern Tariff History, 2d ed. (London: John Murray, 1910), p. 189.Google Scholar

33. Pincus, , Pressure Groups and Antebellum Tariffs, p. 158.Google Scholar

34. Stanwood, , American Tariff Controversies, p. 359.Google Scholar

35. Terrill, , The Tariff, Politics and American Foreign Policy, pp. 211–13.Google Scholar

36. Baack, Bennett D. and Ray, Edward J., “The Political Economy of Tariff Policy: A Case Study of the United States,” Explorations in Economic History 20 (01 1983), pp. 73–93.CrossRefGoogle Scholar

37. Strong group demands are not inconsistent with electoral factors influencing the tariff, unless one takes a strict public good view of group influence, in which industrial or geographic concentration is seen as the chief predictor of success. In this case, an electoral model and an interest group model pull in opposite directions, since the concentration necessary for collective action will usually mean fewer favorable voter interests both in terms of the number of voters and in terms of the number of districts.

38. Mayhew, David R., Congress: The Electoral Connection (New Haven, Conn.: Yale University Press, 1974), p. 17.Google Scholar

39. Turner, Julius and Schneier, Edward, Party and Constituency: Pressures on Congress, revised ed. (Baltimore, Md.: Johns Hopkins University Press, 1970).Google Scholar

40. See Fiorina, Morris P., Representatives, Roll-Calls and Constituencies (Lexington, Mass.: Lexington Books, 1974)Google Scholar. The same issue arises in analyzing votes on tariff bills: the roll-call vote on the McKinley Tariff was a party line split, but this does not mean that ideology or party loyalty dominated constituency interests.

41. For a discussion of electoral safety and homogeneity, see footnote 64 (below).

42. See Mayhew, , Congress, pp. 61–73Google Scholar; and Fenno, Richard F., Homestyle: House Members and Their Districts (Boston: Little, Brown, 1978), p. 28.Google Scholar

43. Sherman, John, Recollections of Forty Years in the House, Senate and Cabinet, vol. 2 (Chicago: Werner, 1895), p. 1085.Google Scholar

44. Kingdon, John W., Congressmen's Voting Decisions, 3d ed. (Ann Arbor: University of Michigan Press, 1989), p. 37.CrossRefGoogle Scholar

45. See Cassing, , McKeown, , and Ochs, , “Political Economy of the Tariff Cycle,” pp. 850–51.Google Scholar

46. See Pincus, , Pressure Groups and Antebellum Tariffs, pp. 107–9.Google Scholar

47. Lavergne, Real P., The Political Economy of U.S. Tariffs (New York: Academic Press, 1983), p. 131.Google Scholar

48. In The Tariff, Politics and American Foreign Policy, p. 138, Terrill notes that “in the 1888 election the Republicans worked vigorously to bring the high tariff gospel to the South,” arguing that it would lead to industrialization.

49. Diminishing marginal utility for a good implies that the price consumers will pay for each successive unit falls. However, the total value of consumption still rises, albeit in diminishing increments, producing a quadratic total revenue curve. The total amount of CHGT accruing to a district is analogous to total revenue.

50. See Peltzman, Sam, “An Economic Interpretation of the History of Congressional Voting in the Twentieth Century,” American Economic Review 75 (09 1985), pp. 656–75.Google Scholar

51. See Fiorina, Morris P., “Some Problems in Studying the Effects of Resource Allocation in Congressional Elections,” American Journal of Political Science 25 (08 1981), pp. 543–67.CrossRefGoogle Scholar

52. Erikson, Robert S., “Is There Such a Thing as a Safe Seat?” Polity 8 (Summer 1974). pp. 623–32.CrossRefGoogle Scholar

53. See Fiorina, , Representatives, Roll-Calls and Constituencies, pp. 121–24Google Scholar; and Kingdon, , Congressmen's Voting Decisions, pp. 29–67Google Scholar. Kingdon associates safety and long tenure with following constituency interests on “high salience votes,” a term that surely fits the McKinley Tariff.

54. Thompson, Joel A. and Moncrief, Gary, “Pursuing the Pork in a State Legislature,” Legislative Studies Quarterly 13 (08 1988), pp. 393–402.CrossRefGoogle Scholar

55. Cox, Gary W., The Efficient Secret (New York: Cambridge University Press, 1987), p. 151.CrossRefGoogle Scholar

56. See footnote 49 (above).

57. The 1880 census was the last one to list employment by county and by industry. The practice of listing this data was subsequently abandoned on the grounds that it revealed confidential information about specific firms. The index uses 1880 employment data in calculating T90 and T83, so that CHGT is holding employment constant. Although this introduces bias into the index, it avoids a worse problem. If CHGT included changes in both employment and tariffs, it would be unclear whether a relationship between CHGT and the electoral variables was due to the political market for tariffs or to changes in employment.

58. See Clausen, Aage R., How Congressmen Decide (New York: St. Martin's Press, 1973), p. 156Google Scholar; and Kingdon, , Congressmen's Voting Decisions.Google Scholar

59. Fiorina, , Representatives, Roll-Calls and Constituencies.Google Scholar

60. District protection (the T_j index) can be broken down into two parts: the part due to the level of manufacturing activity in the district and the part due to the variation in the industry composition of the district. In this case, equation 6 would be as follows:

where MFG_j is the manufacturing employment as a proportion of total population. If a district's change in tariff protection is related more to the across-electorate variation in the overall level of manufacturing industry in the district (MFG_j) than it is to the across-industry variation in the industry composition of the district (e_ij/Σ_ie_ij), then the geographic discrimination in the tariff will be inflexible, always having a greater effect in districts with a high level of manufacturing industry, and only one district could ever have market-clearing tariff rates. This is not the case: the bivariate correlation of CHGT with the across-electorate variation (r = −.024) is insignificant, but the correlation of CHGT with the across-industry variation (measured as CHGT/MFG) is high (r =.549). The association of district protection with differences in the industry composition of districts suggests that the degree of geographic flexibility in the tariff is sufficient to produce tariff rates that will approximately clear each district's political market. Independent evidence supporting my analysis on this point is provided by Peltzman, who notes that the economic heterogeneity of states and regions of the United States (a variable similar to my variation in the industry composition of districts) was high at the beginning of the twentieth century. Brady and Althoff make a similar point in noting the agricultural-industrial “polarization” of the American economy in the period 1890–1910. See Peltzman, , “An Economic Interpretation of the History of Congressional Voting in the Twentieth Century,” p. 659Google Scholar; and Brady, and Althoff, , “Party Voting in the U.S. House of Representatives.”Google Scholar

61. The pseudo-R² used here is that of McKelvey and Zavoina, cited by Aldritch, John H. and Nelson, Forest D. in Linear Probability, Logit and Probil Models (Beverly Hills, Calif.: Sage, 1984), pp. 57–58CrossRefGoogle Scholar. It is the ratio of the explained sum of squares divided by the sum of the explained sum of squares and the number of observations.

62. The regional variables are omitted because of collinearity between party and region.

63. See Stanwood, , American Tariff ControversiesGoogle Scholar; and Taussig, , The Tariff History of the United States.Google Scholar

64. I would note, however, that the data do not support Fiorina's hypothesis about homogeneous districts being safer (see Representatives, Roll-Calls and Constituencies). If this were true, there should be a relationship between V88 and the proportion of manufacturing industry in a district (MFG): V88 = a · MFG² + b · MFG + c, where a > 0. An estimation of this equation, controlling for regional variation, produced an adjusted R² =.17, with a significant coefficient value for a < 0, indicating that heterogeneous districts (medium levels of manufacturing industry) are safest.

65. Terrill, , The Tariff, Politics and American Foreign Policy, p. 11.Google Scholar

66. Mayhew, , Congress, p. 28.Google Scholar

67. Lewis, W. Arthur, Growth and Fluctuations, 1870–1913 (London: Allen & Unwin, 1978), pp. 57–58.Google Scholar

68. See Fiorina, Morris P., “Economic Retrospective Voting in American National Elections,” American Journal of Political Science 22 (05 1978), pp. 426–43CrossRefGoogle Scholar; and Mayhew, , Congress, p. 29.Google Scholar